D
dbeavon
Guest
Are you able to create a timeout scenario and test your code? Maybe you can just change the host name in the url to something that doesn't exist. I'm fairly certain you will catch an error from that code if anything goes wrong. OO programming and SEH error handling work well together, and I'd guess that all these new API's would throw errors by convention. It is much better for API's to follow a standard pattern than come up with their own strange and unique error -publication strategies (just thinking about ugly API's like OS-COPY give me a cold chill). Are you already familiar with structured error handling (SEH)? If not I would highly recommend that you read the "Error Handling" document from beginning to end. community.progress.com/.../2911.openedge-11-7-product-documentation The most important parts in there are related to "Structured Error Handling". This is a similar error-handling pattern to what is found in other languages like Java and C#. I suspect that if you put a do/catch block around all that code, you will probably be able to CATCH a Progress.Lang.Error (or more specifically a class that derives from it). See. documentation.progress.com/.../index.html Have you tried that? The OpenEdge.Net.HTTP stuff is open source so you can go look for yourself to see if there are lots of UNDO, THROW statements in the case of error conditions. This will be your biggest clue. Here is a link to another conversation about that API. Peter Judge seems to be the expert, if not the author: community.progress.com/.../36251
Continue reading...
Continue reading...